Dust is the operating system for AI agents in the enterprise. This model-agnostic platform connects AI models like GPT, Claude, and Gemini to your company's data and workflows without writing code. Create powerful agents in seconds to empower sales, support, marketing, and engineering teams, backed by enterprise-grade security including SOC 2 certification.




If you've ever tried to deploy AI across your organization, you know the frustration. Your data lives in silos—CRM systems, document repositories, Slack channels, GitHub repos. Your teams use dozens of tools that don't talk to each other. And scaling AI beyond a few enthusiastic engineers feels impossible.
That's exactly the problem Dust was built to solve.
Dust is the Operating System for AI Agents—a model-agnostic platform that connects enterprise data sources and workflows to powerful AI models. Think of it as the layer that makes AI actually work in a real business environment, without requiring you to write code or rebuild everything from scratch.
What makes Dust different? Unlike platforms that lock you into a single AI model, Dust lets you choose the best model for each task—OpenAI, Anthropic, Gemini, Mistral, or others. You connect your data sources once, build AI agents that understand your business context, and deploy them across teams in minutes.
The results speak for themselves. Over 2,000 organizations already trust Dust, including companies like Blueground, Clay, Assembled, Patch, Persona, Vanta, Qonto, and Watershed. Clay calls Dust "the most impactful software we've built since constructing Clay itself." Doctolib employees report gaining "20% more time" back in their workday. Back Market built an AI detection system in just one week, contributing €1.2 million in savings.
Whether you're a sales team needing instant customer insights, a support team wanting to cut ticket response from 6 minutes to seconds, or an engineering team looking to automate code reviews—Dust gives your teams AI agents that actually understand your business.
What can Dust actually do for your team? Let's walk through the capabilities that make it happen.
You don't need to be a developer to create powerful AI agents. Dust's no-code builder lets you construct agents that connect to your enterprise data, execute actions, and integrate with your workflows—in just a few seconds. Your team members can build personalized agents for their specific departments without waiting on engineering resources.
Dust integrates directly with the tools your team already uses: GitHub, Google Drive, Notion, Slack, Zendesk, and more. Your data stays private and secure—we never use it to train external models. These connections let your AI agents pull context from across your organization, breaking down information silos that slow everyone down.
Dust supports OpenAI, Anthropic, Gemini, Mistral, and other leading models. You're not locked into a single provider. Switch between models based on performance, cost, or specific capabilities. Need reasoning precision? Use Claude. Want the latest multimodal capabilities? Try Gemini. With Dust, you have flexibility.
Dust is built for team-wide adoption. "Explorers" in your organization can build agents for their teams, gather feedback, and continuously improve. This approach has helped customers like Patch achieve 70% weekly agent usage across team members, and Watershed reach 90% team adoption.
Need something specific? Write custom tools and integrations in code without affecting other functionality. Connect to internal systems, build proprietary agent capabilities, and extend the platform to match your unique workflows.
For complex workflows, Dust lets you build and manage teams of specialized agents. Agents collaborate with each other and with humans, handling sophisticated automation scenarios that would otherwise require significant engineering effort.
Security isn't an afterthought—it's foundational. Dust is SOC 2 Type II certified, GDPR compliant, and supports HIPAA requirements. Your data is encrypted at rest (AES-256) and in transit (TLS). You get granular access controls, SSO/SCIM support, and data residency options in the US or EU.
Dust serves teams across every department. Here's how different functions are putting AI agents to work—and the results they're seeing.
The challenge: Manually compiling customer information, spending hours preparing for outreach.
The solution: Create account snapshots from CRM data in seconds. Use call recordings to generate personalized outreach. Automate RFP responses that used to take days.
The impact: Sales reps get instant answers to prospect questions, freeing them to focus on closing deals.
The challenge: Slow response times, knowledge scattered across wikis and team members.
The solution: Connect agents to your knowledge base for instant, accurate responses. Agents identify ticket patterns and route issues appropriately.
The impact: Malt reduced average ticket handling time from 6 minutes to just seconds. Support teams handle more volume without adding headcount.
The challenge: Content creation eats time, maintaining brand consistency across channels is difficult.
The solution: Draft brand-aligned content in minutes. Create consistent messaging for campaigns across channels.
The impact: Brevo achieved 80% efficiency improvement in personalized content creation and eliminated engineering tickets for marketing needs.
The challenge: Debugging takes hours, documentation falls out of date, code reviews bottleneck.
The solution: Use code context directly in your IDE—documentation, issue history, and more—to accelerate debugging. Automate code reviews and generate documentation updates.
The impact: Engineers spend less time on repetitive tasks and more time on what matters: building great products.
The challenge: Contract review is time-consuming, compliance tracking is manual.
The solution: Agents review contracts and flag potential issues. Generate standard legal documents from templates.
The impact: Didomi dramatically improved privacy review efficiency—handling compliance checks, suggesting improvements, and drafting communications.
The challenge: Information lives in silos, searching across systems is painful.
The solution: Build agents that access company information across all data silos instantly. Create searchable knowledge bases that actually work.
The impact: Kyriba's agents became "knowledge partners" for all employees, dramatically improving productivity.
The challenge: Ticket queues pile up, common issues consume too much time.
The solution: First-line technical support agents that handle common IT questions and guide users through routine processes.
The impact: IT teams focus on complex issues while agents handle the basics.
The challenge: Onboarding is cumbersome, policy questions consume HR bandwidth.
The solution: Agents guide new employees through onboarding and answer HR policy and benefits questions.
The impact: HR teams reclaim time for strategic work while employees get instant answers.
Select the department that matches your role to see how Dust solves your specific challenges. Each team experiences different pain points—and Dust tailors accordingly.
Curious about how Dust works under the hood? Here's a deeper look at the technology that makes it possible.
Dust is built on "operating system primitives"—fundamental architectural concepts that let AI models connect to your enterprise data and tools, much like an OS connects applications to hardware. This isn't a thin wrapper around an LLM. It's a purpose-built infrastructure for enterprise AI.
Unlike platforms that push you toward their own models, Dust's architecture is genuinely model-agnostic. You can use OpenAI's latest models, Anthropic's Claude family, Google's Gemini, Mistral's open models, or others—switching is as simple as a configuration change. This protects you from vendor lock-in and lets you adopt improvements from any provider.
Dust agents can invoke multiple tools in sequence: semantic search across your documents, data analysis on connected datasets, web navigation for real-time information, and custom actions. These tools work together to give agents the context they need to solve complex problems.
Official integrations include GitHub, Google Drive, Notion, Slack, Zendesk, Salesforce, and more. Beyond these, you can build custom integrations using Dust's API. Your data connections are configured once and reused across all your agents.
Every layer of Dust is designed with security in mind:
Dust offers two tiers designed for different organization sizes and needs.
Perfect for small to medium teams getting started with AI agents. Starting at 1 user with a 14-day free trial—no credit card required.
What's included:
For large organizations needing advanced security, control, and support. Includes everything in Pro plus:
| Feature | Pro | Enterprise |
|---|---|---|
| Price | €29/user/month | Custom (100+ users) |
| Users | 1+ | 100+ |
| Advanced models | ✓ | ✓ |
| Custom agents | ✓ | ✓ |
| Data connectors | 7+ | Unlimited |
| SSO | — | ✓ |
| Priority support | — | ✓ |
| US/EU hosting | — | ✓ |
| SCIM | — | ✓ |
For teams of 1-20, the Pro plan provides everything you need to get started. If you're scaling beyond 100 users or need advanced security features like SSO, the Enterprise plan offers the controls and support that larger organizations require.
Dust is a model-agnostic operating system for AI agents. While many platforms focus on training larger models, Dust focuses on connecting AI models to your enterprise data and tools—creating agents that actually work in your business context.
Absolutely. Your data is encrypted at rest (AES-256) and in transit (TLS 1.3). Dust is SOC 2 Type II certified, GDPR compliant, and supports HIPAA requirements. We offer zero data retention—your data stays yours and is never used to train external models.
Dust supports all major models: OpenAI (GPT-4/5), Anthropic (Claude), Google (Gemini), Mistral, and others. You choose the best model for each task and can switch anytime.
Pro plan is €29 per user per month (excluding VAT). Enterprise plans are custom-priced for organizations with 100+ users. Both include a 14-day free trial—no credit card required to start.
Sign up for a 14-day free trial at dust.tt. No credit card needed. You can start building agents immediately and connect your data sources in minutes.
Dust integrates with GitHub, Google Drive, Notion, Slack, Zendesk, Salesforce, and more. You can also build custom integrations using our API for internal systems.
Yes, Enterprise plans include SSO support for Okta, Microsoft Entra ID, and Jumpcloud. SCIM for automated user provisioning is also available.
Dust is the operating system for AI agents in the enterprise. This model-agnostic platform connects AI models like GPT, Claude, and Gemini to your company's data and workflows without writing code. Create powerful agents in seconds to empower sales, support, marketing, and engineering teams, backed by enterprise-grade security including SOC 2 certification.
One app. Your entire coaching business
AI-powered website builder for everyone
AI dating photos that actually get matches
Popular AI tools directory for discovery and promotion
Product launch platform for founders with SEO backlinks
Cursor vs Windsurf vs GitHub Copilot — we compare features, pricing, AI models, and real-world performance to help you pick the best AI code editor in 2026.
Compare the top AI agent frameworks including LangGraph, CrewAI, AutoGen, OpenAI Agents SDK, and LlamaIndex. Find the best framework for building multi-agent AI systems.